Recognition method for apple fruit based on SUSAN and PCNN

Xu, Liming; Lv, Jidong

doi:10.1007/s11042-017-4629-6

Cited by 28 publications

(20 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here we provide some experiments to verify the superiority of our algorithm, comparing with several state-of-the-art approaches utilizing quantum-behaved particle swarm optimization (QPSO) for muti-threshold searching [20], fuzzy c-means clustering (FCM) [16], pulse coupled neural network-based segmentation (PCNN) [14], Co-segmentation [53] and the traditional LDA, seen from Figures 17-20. We use red, green, and blue color to represent the labeled area of fruits, leaves, and the background, respectively.…”

Section: Comparison Settings and Resultsmentioning

confidence: 99%

“…We selected three unsupervised segmentation algorithms [20], [16], and [14] in the field of plant segmentation for comparative experiments. Through the results, we can see that when the same segmentation algorithm is applied to diverse backgrounds, their weaknesses will emerge at different level.…”

Section: Analysis Of Comparative Experimentsmentioning

confidence: 99%

“…Artificial pulse coupled neural network PCNN is used in [14] for the recognition and segmentation of fruits. It matches the pixels with neurons and connects each other to form a single-layer, two-dimensional local feedback network.…”

Section: Analysis Of Comparative Experimentsmentioning

confidence: 99%

“…It retains more information from the original image with the lowest similarity between classes and the greatest one within classes by the objective function. Although the concept of neighborhood density is introduced to initialize the clustering center in the literature and it does show higher segmentation accuracy than the above two, however, FCM is effective only for the recognition of mature tomatoes in close shots similar to [14]. As shown in image 4,7,8,10,17,19, and 20, all tomatoes have very high visual saliency in the images.…”

Section: Analysis Of Comparative Experimentsmentioning

confidence: 99%

“…Until now, a diverse range of image segmentation algorithms have been presented in the agriculture field, among which there are color-based threshold methods [10,11], active contour methods combined with prior knowledge [12,13], neural network-based methods [14], and graph-clustering methods [15][16][17]. In order to obtain more accurate results, supervised algorithms [18,19] are also involved.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Unsupervised Greenhouse Tomato Plant Segmentation Based on Self-Adaptive Iterative Latent Dirichlet Allocation from Surveillance Camera

Cao

2019

Agronomy

View full text Add to dashboard Cite

It has long been a great concern in deep learning that we lack massive data for high-precision training sets, especially in the agriculture field. Plants in images captured in greenhouses, from a distance or up close, not only have various morphological structures but also can have a busy background, leading to huge challenges in labeling and segmentation. This article proposes an unsupervised statistical algorithm SAI-LDA (self-adaptive iterative latent Dirichlet allocation) to segment greenhouse tomato images from a field surveillance camera automatically, borrowing the language model LDA. Hierarchical wavelet features with an overlapping grid word document design and a modified density-based method quick-shift are adopted, respectively, according to different kinds of images, which are classified by specific proportions between fruits, leaves, and the background. We also utilize the feature correlation between several layers of the image to make further optimization through three rounds of iteration of LDA, with updated documents to achieve finer segmentation. Experiment results show that our method can automatically label the organs of the greenhouse plant under complex circumstances, fast and precisely, overcoming the difficulty of inferior real-time image quality caused by a surveillance camera, and thus obtain large amounts of valuable training sets.

show abstract